Costco Stock Forecast: How to Read an Analyst Price Target Properly

Type the words Costco stock forecast into a search box and one figure comes back, the averaged analyst target, usually with a percentage gap to today's quote bolted on. It looks like a prediction. It is an average of opinions, each built from assumptions you can check and most of which are published by the company itself.

MarketBeat's page on the company, read on 12 October 2026, showed an average twelve-month forecast of $1,046.27 drawn from 33 analyst ratings, with a high of $1,315.00 and a low of $16.00, next to a quoted price of $958.65 and a consensus label of Moderate Buy. Hold onto that low figure. It matters later.

What Sits Inside a Target

A price target is two forecasts multiplied together. The first is earnings per share for some future period, usually the next twelve months. The second is the multiple the analyst thinks the market will pay for those earnings. Multiply and you have a share price.

Both halves are judgements. Earnings can be modelled line by line from the company's disclosures, and reasonable people land close to each other. The multiple is a guess about what other investors will feel like paying, and it is where almost all the disagreement between analysts lives. A target is therefore mostly a statement about sentiment wearing the clothes of a calculation.

Which is why the right way to use one is backwards: work out what earnings and what multiple the number implies, then decide whether you find either assumption plausible. The broader case for reading targets this way sits in what a price target is and how accurate they turn out to be.

The Inputs You Can Verify Without Anyone's Help

For a warehouse club the earnings half of the calculation runs on four things, and the company reports all of them. Its results release for the 52 weeks to 30 August 2026 gives the lot.

• Comparable sales, meaning growth from warehouses already open. The release reports total company comparable sales of 8.4 per cent for the year, or 6.6 per cent after stripping out petrol prices and currency, with digitally-enabled comparable sales of 20.9 per cent.

• New openings, which add sales without touching the comparable figure. The company operated 939 warehouses at the year end, 647 of them in the United States and Puerto Rico, and MarketBeat's write-up of the results reported a plan for 33 warehouse openings during fiscal 2027, five of which are relocations, giving 28 net new buildings.

• Membership income, which is close to pure profit. The release puts membership fees at $5.907 billion for the year against operating income of $11.685 billion, so roughly half of operating profit arrived as subscription revenue.

• Renewal and member growth, which decide whether that subscription line keeps compounding. MarketBeat reported paid members up 3.8 per cent to 84.1 million, executive members up 9.4 per cent to 42.3 million, and a United States and Canada renewal rate of 92.3 per cent.

Four numbers, all published, all free. Anyone who wants to disagree with a consensus estimate has to disagree with one of those four, and saying which one is the difference between a view and a mood. The first of them deserves its own treatment, which is why comparable sales is the number to read first at a retailer.

Separating the One-Off From the Run Rate

Forecasts break most often on items that happened once. Costco's release states that the fourth quarter was helped to the tune of 15 cents per diluted share by a non-recurring tariff refund received in the quarter, partly offset by reinvesting those refunds in member value. Full-year diluted earnings came in at $20.76 a share against $18.21 the year before.

Fifteen cents inside a quarter that earned $6.75 a share is not large, and the company flagged it in the same breath as the result. The point is procedural rather than arithmetic. Any item the company calls non-recurring has to come out before you annualise anything, and management said it expects to reinvest most future refunds in lower member prices, which means the benefit is not a run-rate item at all.

What the Rulebook Demands, and What the Free Page Shows

This is the part almost no forecast article covers. Analyst research in the United States is regulated, and the rules are specific about what has to travel with a price target.

FINRA Rule 2241 requires that any recommendation, rating or price target has a reasonable basis and is accompanied by a clear explanation of any valuation method used and a fair presentation of the risks that may impede achievement of the target. The same rule requires a firm's rating system to be defined in each report, including the time horizon and any benchmark the rating is measured against.

It goes further. A firm must state the percentage of all the securities it rates to which it would assign a buy, hold or sell rating, and the percentage of companies in each of those categories for which it has provided investment banking services in the previous twelve months. Where a rating or target has been in place for at least a year, the report must carry a price graph of the security.

Now compare that with the thing you can actually see. The free aggregator page carries a number, a count of analysts and a label. The valuation method, the stated time horizon, the risks, the firm's own distribution of ratings and the banking relationships are all left behind in a document retail investors never see. Not one of those disclosures survives aggregation.

A Consensus Average Is Only as Clean as Its Worst Entry

The position, and its evidence sits in the opening paragraph above. That MarketBeat page listed a low forecast of $16.00 alongside a high of $1,315.00 and an average of $1,046.27 across 33 ratings. A $16.00 twelve-month target on a company earning more than twenty dollars a share is not an opinion. It is almost certainly a stale entry, a data error, or a figure that was never adjusted for a share split. It is also being quietly carried in the same list that produces the average everybody quotes.

One bad entry barely moves an average of 33. That is not the problem. The problem is what its presence tells you about the dataset: nothing in the pipeline checks whether an individual target still means what it meant when it was filed, and the regulated disclosures that would let you judge any single entry have been stripped out. A number assembled that way is a sentiment indicator with a decimal point, and it should be read with the same seriousness as any other crowd measure.

So the practical rule is to use the spread rather than the average, and to use it as a question rather than an answer. A wide range means the disagreement is about the multiple, which tells you the argument is over what investors will pay rather than over what the company will earn. Then go and form your own view from the four published inputs above, which is the method behind working through a valuation on this company from scratch.

Consensus targets assembled with the same discipline as the underlying research, with stale entries dropped and methods attached, would be a different matter. If a provider published the time horizon and the valuation basis for every target in its average, the average would deserve more weight than the spread. No free page does that today.

What Would Change the Estimates

Three things move this particular set of forecasts. Comparable sales excluding petrol and currency, because that is the cleanest read on whether shoppers are spending more. The renewal rate, because a fall there threatens the profit line that justifies the whole valuation. And the multiple itself, which is not a company event at all, and taken up in why this particular retailer trades where it does on earnings.

The Short Version

A forecast is a bundle of judgements with a price stapled to the front. Unpick the judgements and the business comes into focus. Read only the staple and you have borrowed a stranger's conclusion without the paperwork that would let you weigh it.

This article is for educational and informational purposes only. It is not financial advice, and Investor Muse is not a registered investment advisor. Historical figures describe the past and do not predict future results. Speak with a qualified professional before you make decisions about your money.

Position disclosure: neither Investor Muse nor its editorial team holds a position in any company, fund or security named in this article.